Since conventional power supplies mounted on high frequency heating apparatus have been made heavy and bulky, these conventional power supplies are desired to be made compact and light weight. As a result, various technical ideas capable of manufacturing these power supplies in compact, light weight and low cost have be actively progressed in such a way that these power supplies are constructed in switching modes. In high frequency heating apparatus which cook food products by using microwaves generated by magnetrons, various needs capable of making power supplies compact and in light weight have been requested, which are employed so as to drive magnetrons. These needs could be realized by switching-type inverter circuits.
Among these switching type inverter circuits, more specifically, a high frequency inverter circuit which is directed by the present invention corresponds to a resonant type circuit system with employment of two switching elements which construct arms of bridge circuits (refer to, for example, JP-A-2000-58252).
When a 1-transistor type inverter (width ON/OFF-control type inverter) is arranged, a withstanding voltage between a collector and an emitter of this transistor requires approximately 1000V. However, when a 2-transistor type inverter having abridge circuit is arranged, withstanding voltages between collectors and emitters of these transistors are not required to be so high withstanding voltages. As a result, if the inverter circuit is constructed of the bridge circuit arrangement, then the withstanding voltages between the collectors and the emitters of these transistors may be lowered to approximately 600 V. Accordingly, there is such a merit that low-cost transistors may be used in these transistor inverters. In this sort of inverter, while a resonant circuit is constituted by an inductance “L” and a capacitance “C”, this inverter owns such a resonance characteristic as represented in FIG. 1, in which a resonant frequency “f0” is defined as a peak.
FIG. 1 is a graphic diagram for representing a current-to-used frequency characteristic in the case that a constant voltage is applied to an inverter resonant circuit according to the present invention.
A frequency “f0” corresponds to a resonant frequency of an LC resonant circuit of the inverter circuit, and a current-to-frequency characteristic curve “I1” of a frequency range defined from “f1” to “f3”, which is higher than this resonant frequency “f0” is utilized.
At the resonant frequency “f0”, a current I1 becomes maximum, and in connection with an increase of the frequency range from f1 to f3, this current I1 is decreased. Within the frequency range defined from f1 to f3, the lower the frequency is decreased, the closer the frequency is approached to the resonant frequency f0, so that the current I1 is increased. As a result, a current flowing through a secondary winding of a leakage transformer is increased. Conversely, the higher the frequency is increased, the further the frequency is separated apart from the resonant frequency f0, so that the current I1 is decreased. As a result, the current flowing through the secondary winding of the leakage transformer is decreased. In the inverter circuit for driving a microwave oven which functions as a non-linear load, since this frequency is varied, power of the microwave oven is changed.
As will be explained later, in the case that an input power supply for a microwave oven using a non-linear load of a magnetron corresponds to an AC source such as a commercial power supply, the microwave oven causes a switching frequency to be changed.
As to respective high frequency power of a microwave oven, the highest frequency appears at temperatures of approximately 90 degrees and about 270 degrees. For instance, when the microwave oven is operated at 200 W, the operating frequency is approached to f3; when the microwave oven is operated at 500 W, the operating frequency is lower than f3; and when the microwave oven is operated at 1000 W, the operating frequency is further lower than f3. Apparently, since either an input power control or an input current control is carried out, this frequency may be changed in accordance with changes as to voltages of a commercial power supply, temperatures of the microwave oven, and the like.
Also, in the vicinity of 0 degree and 180 degrees of phases of the above-described power supply voltage, since the operating frequency of the magnetron is set near the frequency “f1” which is close to the resonant frequency “f0” where the resonant current is increased in correspondence with such a characteristic of a magnetron that if a high voltage is not applied thereto, then this magnetron is not resonated in a high frequency, a boosting ratio of the voltage applied to the magnetron to the voltage of the commercial power supply is increased, and also, the phase width of the commercial power supply is set to be widened, by which electromagnetic waves are produced from the magnetron.
FIG. 2 shows an example of a resonant type high frequency heating apparatus operated by switching elements of a two-element bridge circuit, which is described in JP-A-2000-58252. In FIG. 2, the high frequency heating apparatus has been arranged by a DC power supply 1, a leakage trans former 2, a first semiconductor switching element 6, a first capacitor 4, a second capacitor 5, a third capacitor (smoothing capacitor) 13, a second semiconductor switching element 7, a driving unit 8, a full-wave doubler rectifying circuit 10, and a magnetron 11.
The DC power supply 1 rectifies an AC voltage of a commercial power supply in a full-wave rectification mode to produce a DC voltage VDC, and then, applies the DC voltage VDC to a series circuit constituted by the second capacitor 5 and a primary winding 3 of the leakage transformer 2. While the first semiconductor switching element 6 has been series-connected to the second semiconductor switching element 7, the series circuit constituted by the primary winding 3 of the leakage transformer 2 and the second capacitor 5 has been connected parallel to the second semiconductor switching element 7.
The first capacitor 4 has been connected parallel to the second semiconductor switching element 7. An AC high voltage output generated from a secondary winding 9 of the leakage transformer 2 has been converted into a DC high voltage by the full-wave doubler rectifying circuit 10, and then, this DC high voltage has been applied between an anode and a cathode of the magnetron 11. A thirdly winding 12 of the leakage transformer 2 supplies a current to the cathode of the magnetron 11.
The first semiconductor switching element 6 has been constituted by an IGBT (Insulated Gate Bipolar Transistor) and a flywheel diode connected parallel to the IGBT. Similarly, the second semiconductor switching element 7 has been constituted by an IGBT and a flywheel diode connected parallel to the IGBT.
As apparent from the foregoing description, both the first and second semiconductor switching elements 6 and 7 are not limited only to the above-explained sort of semiconductor switching element, but a thyristor, a GTO (Gate Turn Off) switching element, and the like may be alternatively employed.
The driving unit 8 contains an oscillating unit which is used so as to produce drive signals for driving the first semiconductor switching element 6 and the second semiconductor switching element 7. While this oscillating unit oscillates the drive signals having predetermined frequencies and duty ratios, the driving unit 8 has applied these drive signals to the first semiconductor switching element 6 and the second semiconductor switching element 7.
The first semiconductor switching element 6 and the second semiconductor switching element 7 are alternately driven, or are driven by providing such a time period during which both the first and second semiconductor switching elements 6 and 7 are commonly turned OFF, namely by providing a dead time by employing a dead time forming means (will be explained later). Just after any one of the first and second semiconductor switching elements 6 and 7 has been turned OFF, a voltage across the terminals of the other semiconductor switching element is high. As a result, if the other semiconductor switching element is turned ON at this time, then an excessively large current having a spike shape may flow through this turned-ON switching element, so that unwanted loss and undesirable noise may be produced. However, since this turn-ON operation may be delayed until the high voltage across the switching element is decreased to approximately 0 V, the above-described loss and noise may be prevented. Apparently, a similar operation may be carried out when the switching element opposite to the above-described switching element is turned OFF.
FIG. 3 indicates respective modes in which the circuit of FIG. 2 is operated.
Also, FIG. 4 shows a voltage and current waveform diagram as to components such as semiconductor switching elements employed in the circuit.
In the drawing, in a mode 1 of FIG. 3(a), a drive signal is supplied to the first semiconductor switching element 6. At this time, a current flows from the DC power supply 1 through both the primary winding 3 of the leakage transformer 2 and the second capacitor 5.
In a mode 2 of FIG. 3(b), the first semiconductor switching element 6 is turned OFF, and the current which has flown through the primary winding 3 and the second capacitor 5 starts to flow along a direction to the first capacitor 4, and at the same time, the voltage of the first semiconductor switching element 6 is increased.
In a mode 3 of FIG. 3(c), the voltage of the first capacitor 4 is directed from VDC to zero V. In the mode 3, the voltage across both the terminals of the first capacitor 4 is reached to zero V, so that the diode which constitutes the second switching element 7 is turned ON.
In a mode 4 of FIG. 3(d), since the direction of the current is inverted which has flown through the primary winding 3 and the second capacitor 5 due to the resonant phenomenon, at this time, the second semiconductor switching element 7 must be turned OFF. In the time periods of the modes 2, 3, and 4, the voltage of the first semiconductor switching element 6 becomes equivalent to the DC power supply voltage VDC. In such a region as Europe where an effective value as to a commercial power supply voltage is 230 V, since a voltage peak becomes root 2 times higher than the effective voltage, a DC power supply voltage VDC becomes near equal to 325 V.
In a mode 5 of FIG. 3(e), the second semiconductor switching element 7 is turned OFF, and the current which has flown through the second capacitor 5 and the primary winding 3 starts to flow along a direction to the first capacitor 4, so that the voltage of the first capacitor 4 is increased up to the VDC.
In a mode 6 of FIG. 3(f), the voltage of the first capacitor 4 is reached to the voltage VDC, and thus, the diode which constitutes the first semiconductor switching element 6. Since the direction of the current is inverted which has flown through the primary winding 3 and the second capacitor 5 due to the resonant phenomenon, at this time, the first semiconductor switching element 6 must be turned ON, which constitutes the mode 1. In the time periods of the modes 1 and 6, the voltage of the second semiconductor switching element 7 becomes equivalent to the DC power supply voltage VDC.
In accordance with this circuit arrangement, a maximum value as to the voltages applied to both the first semiconductor switching element 6 and the second semiconductor switching element 7 can be set to the DC power supply voltage VDC.
Both the mode 2 and the mode 5 correspond to such a resonant period during which the current flown from the primary winding 3 may flow through the first capacitor 4 and the second capacitor 5. Since a capacitance value of the first capacitor 4 has been set lower than, or equal to 1/10 of a capacitance value of the second capacitor 5, a combined capacitance value becomes nearly equal to the capacitance value of the first capacitor 4. The voltages applied to the first semiconductor switching element 6 and the second semiconductor switching element 7 in the modes 3 and 5 are changed based upon a time constant which is determined by this combined capacitance value and an impedance of the leakage transformer 3. Since this voltage change owns such an inclination which is determined based upon the above-explained time constant, the switching loss occurred when the first semiconductor switching element 6 is turned OFF in the mode 3 may be reduced.
Moreover, in the mode 5, since the voltage becomes zero, when the first semiconductor switching element 6 is turned ON in the mode 1, the voltage applied to the first semiconductor switching element 6 becomes zero, so that the switching loss of the first semiconductor switching element 6 can be reduced when this switching element 6 is turned ON. This is referred to as a “zero voltage switching” operation, and these items are features of the resonant circuit system. The present system utilizes these features, and owns such a merit that a voltage of a semiconductor switching element does not become higher than, or equal to the DC power supply voltage VDC. As shown in FIG. 4, the capacitance value of the second capacitor 5 has been set to a sufficiently large capacitance value in such a manner that the voltage of this second capacitor 5 contains a small ripple component.
On the other hand, as shown in FIG. 2, in such an inverter circuit that the series connection circuit constructed of the first and second switching elements 6 and 7 has been connected parallel to the DC power supply 1 and the arm has been constituted by two switching elements, since the ON/OFF operations of the first and second semiconductor switching elements 6 and 7 are alternately repeated, the high frequency AC voltage is generated in the primary winding 3 of the leakage transformer 2, and then, the high frequency high-voltage is induced in the secondary winding 9. Such an instantaneous time period during which both the first and second semiconductor switching elements 6 and 7 are turned ON at the same time is not completely provided. This is because the shortcircuit of the DC power supply 1 may occur.
Under such a circumstance, conventionally, a time period (will be referred to as “dead time” and will be abbreviated as “DT”) has been necessarily provided, during which both the first and second switching elements 6 and 7 are not turned ON after any one of the first and second semiconductor switching elements 6 and 7 has been turned OFF until the remaining semiconductor switching element is turned ON.
Now, the dead time (DT) will be explained with reference to FIG. 4.
FIG. 4 indicates voltage waveforms and current waveforms as to the first and second semiconductor switching elements 6 and 7 (FIG. 2), and the first and second capacitors 4 and 5 (FIG. 2) in the above-explained respective modes 1 to 6.
A part (a) of FIG. 4 shows a current waveform of the first semiconductor switching element 6 in the above-explained respective modes 1 to 6. The first semiconductor switching element 6 which had been conducted from a time instant “t0” (accordingly, voltage between emitter and collector of first semiconductor switching element 6 becomes zero in (b) of FIG. 4) has been turned OFF (namely current becomes zero) at an ending time instant “t1” of the mode 1.
On the other hand, (d) of FIG. 4 shows a voltage waveform of the second semiconductor switching element 7. The second semiconductor switching element 7 which has been turned OFF from the time instant “t0” is continued to be turned OFF until a starting time instant “t2” of the mode 3 in which an ON signal is applied.
As a consequence, in a time period “DT1” defined from the time instant “t1” up to the time instant “t2”, both the first semiconductor switching element 6 and the second semiconductor switching element 7 are commonly turned OFF.
This time period DT1 corresponds to a minimum value which is required for the dead time. A maximum value of the dead time corresponds to a time period defined from the time instant t1 up to the time instant t3. Thus, the dead time is allowed within this time range.
Similarly, such a time period “DT2” corresponds to a minimum value which is required for the dead time. This time period “DT2” is defined by that after the second semiconductor switching element 7 is turned OFF (namely, current becomes zero) at a time instant “t4” (see (c) of FIG. 4), until an ON signal is applied to the first semiconductor switching element 6 at a starting time instant “t5” of the mode 6 as represented in (a) of FIG. 4. A maximum value of the dead time corresponds to such a time period from the time instant “t4” up to a time instant “t6”. Thus, the dead time is allowed within this time range.
In the conventional 2-transistor type inverter circuit, these dead times “DT” have been defined as the time period “DT1” and the time period “DT2” in such a manner that such a time range is calculated where the turn-ON and turn-OFF operations of the first semiconductor switching element 6 are not overlapped with those of the second semiconductor switching element 7. These time periods DT1 and DT2 have been calculated as fixed values.
In the case of an inverter circuit for a microwave oven, however, when the inverter circuit is driven in a range of a high switching frequency, a time duration after one semiconductor switching element is turned OFF until a voltage Vce between an emitter and a collector of another semiconductor switching element is decreased to 0 V is prolonged. As a consequence, after one semiconductor switching element has been turned OFF, and the fixed dead time has passed, if a turn-ON signal is applied to the other switching element, then the other semiconductor switching element is turned ON while the voltage Vce between the emitter and the collector is not decreased to 0 V. When the switching frequency is high, a heat loss may be produced in the semiconductor switching elements. Thus, malfunction of the semiconductor switching elements and an occurrence of a spike current may constitute a noise generating source.
The reason why the above-explained heat loss and noise are produced will now be explained with reference also to FIG. 4.
That is, even when the first semiconductor switching element 6 is turned OFF (namely, current becomes zero) in the time instant t1 (see (a) of FIG. 4), such a time duration defined by subtracting the time instant “t1” from the time instant “t2” is required in order that the voltage (solid line) across both the terminals of the second semiconductor switching element 7 is dropped to 0 V (see (d) of FIG. 4). As a result, when the turn-ON signal is applied to the other semiconductor switching element 7 at the time instant t2, since the voltage Vce between the emitter and the collector of the second switching element 7 has been decreased to 0 V, this second semiconductor switching element 7 is turned ON (becomes conductive) from the voltage of 0 V (this operation is referred to as “zero volt switching” operation). As a consequence, there is no problem as to the heat loss and the noise.
However, an inclination of a trapezoid of the voltage VDC is changed in response to a strength of resonance. If the resonance is strong (namely, frequency is low), then the inclination becomes sharp, so that the voltage across both the terminals of the first semiconductor switching element 7 quickly becomes zero volt. If the resonance is weak (namely, frequency is high), then the inclination becomes gentle, so that a lengthy time is required in order that the voltage across both the terminals of the first semiconductor switching element 7 is lowered to zero volt.
As previously explained, when the inverter circuit is operated in the high frequency range, the switching frequency is separated apart from the resonant frequency, so that the time constant is prolonged, and in (d) of FIG. 4, a time duration becomes long during which the voltage (indicated by dot line) across both the terminals of the other (second) semiconductor switching element 7 is decreased to 0 V. Thus, this voltage cannot be completely decreased to 0 V during a time period between the time instant t1 and the time instant t2, but even after the time instant t2 has elapsed, a predetermined voltage (refer to symbol Vt2 of dot line F in (d) of FIG. 4) is still applied to this second semiconductor switching element 7.
As a consequence, when the turn-ON signal is applied to the second semiconductor switching element 7 at the time instant “t2” in accordance with the normal operation manner, this second semiconductor switching element 7 is turned ON while the predetermined voltage Vt2 is being still applied between the emitter and the collector of this second semiconductor switching element 7, so that the heat loss has been produced. Also, the steep spike current may flow due to an occurrence of a large dv/dt, which causes the noise source.
Even when such a hard switching operation (namely, switching operation is forcibly carried out even when either voltage or current is not zero) is carried out, since the dead time is secured, this hard switching operation never conducts such a failure that the power supply is shortcircuited, but an extra heat loss may be merely produced in the IGBT. However, since these heat losses are cooled by a heat sink, even when such heat losses may occur, the inverter operation could be continuously carried out under the normal condition. Also, the noise caused by the spike current could not become a considerably large noise value as a serious problem. Accordingly, in the conventional inverter circuits, the failure as to the above-described hard switching operation does not completely constitute the problem. As a consequence, the below-mentioned inverter circuit may be conceived. That is, while the dead time “DT” which has been conventionally fixed is variable, in this inverter circuit, useless energy is not consumed, no adverse influence is given to lifetimes of semiconductor switching elements, and moreover, noise can be hardly produced.
In such a high frequency heating apparatus equipped with a dead time forming circuit which forms either a fixed dead time or a variable dead time, a predetermined time duration after this high frequency heating apparatus has been initiated and until an oscillation of a magnetron becomes stable is required.
Furthermore, when the oscillation of the magnetron is stabilized and is then transferred to the normal condition, such an overshoot phenomenon may occur that an input current Iin flowing through a main circuit exceeds a specified current value. Then, if the input current Iin of the main circuit overshoots, then there is a risk that both the IGBTs and the magnetron are damaged. Therefore, it is rapidly required to suppress the overshoot phenomenon so as to protect these electronic components.